Research Article Atherosclerotic Plaque Stability Is Affected by the Chemokine CXCL10 in Both Mice and Humans

Transcription

1 SAGE-Hindawi Access to Research International Journal of Inflammation Volume 211, Article ID 93619, 9 pages doi:1.61/211/93619 Research Article Atherosclerotic Plaque Stability Is Affected by the Chemokine CXCL1 in Both Mice and Humans Dolf Segers, 1 Jonathan A. Lipton, 1 Pieter J. M. Leenen, 2 Caroline Cheng, 1 Dennie Tempel, 1 Gerard Pasterkamp, 3 Frans L. Moll, Rini de Crom, 5 and Rob Krams 6 1 Department of Cardiology, Erasmus University Medical Center, 3522ZZ Rotterdam, The Netherlands 2 Department of Immunology, Erasmus University Medical Center, 3522ZZ Rotterdam, The Netherlands 3 Department of Cardiology, University Medical Center-Utrecht, 358CX Utrecht, The Netherlands Department of Vascular Surgery, University Medical Center-Utrecht, 358CX Utrecht, The Netherlands 5 Department of Cell Biology and Genetics, Erasmus University Medical Center, 3522ZZ Rotterdam, The Netherlands 6 Department of Bioengineering, Royal School of Mines, Imperial College, London SW7 2AZ, UK Correspondence should be addressed to Rob Krams, r.krams@imperial.ac.uk Received 29 June 211; Accepted 25 August 211 Academic Editor: Elena Aikawa Copyright 211 Dolf Segers et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background. The chemokine CXCL1 is specifically upregulated during experimental development of plaque with an unstable phenotype. In this study we evaluated the functional consequences of these findings in mice and humans. Methods and Results. In ApoE / mice, we induced unstable plaque with using a flow-altering device around the carotid artery. From week 1 to, mice were injected with a neutralizing CXCL1 antibody. After 9 weeks, CXCL1 inhibition resulted in a more stable plaque phenotype: collagen increased by 58% (P =.2), smooth muscle cell content increased 2-fold (P =.3), while macrophage MHC class II expression decreased by 5% (P =.5). Also, the size of necrotic cores decreased by 1% (P =.1). In 16 human carotid endarterectomy specimens we found that increasing concentrations of CXCL1 strongly associate with an increase in atheromatous plaque phenotype (ANOVA, P =.3), with high macrophage, low smooth muscle cell, and low collagen content. Conclusions. In the present study we showed that CXCL1 is associated with the development of vulnerable plaque in human and mice. We conclude that CXCL1 might provide a new lead towards plaque-stabilizing therapy. 1. Introduction Atherosclerosis is a progressive inflammatory disease of the arterial vasculature with lesions that may turn unstable, which may lead to lesion rupture. The high mortality of these ruptures has been associated with atherosclerotic plaques that display a specific histological phenotype, characterized by a large pool of lipids and necrotic cell debris covered by a thin fibrous cap and the presence of inflammatory cells in the plaque shoulders [1]. Previously, we developed an experimental animal model in which plaques with stable and unstable characteristics are simultaneously induced in a single straight arterial segment [2]. We subsequently identified specific expression profiles of various chemokine genes in vulnerable plaque development [3]. In these, the chemokine CXCL1 was specifically upregulated during the early phase in the development of the vulnerable plaque segment. CXCL1 is a member of the CXC family of chemokines. Upon stimulation with interferon-gamma (IFN-γ) it is expressed by many cell types, for example, monocytes/ macrophages, endothelial cells, fibroblasts, and natural killer cells [, 5]. Effects of CXCL1 are mediated by the G- protein coupled receptor CXCR3, which is known to be expressed by resting and activated T cells, NK cells, and a subset of peripheral blood monocytes [5 7]. Furthermore, CXCL1 plays a variety of roles in inflammatory diseases, like in the initiation and maintenance of T-helper-1- (Th1-) polarized immune responses [8, 9], migration of activated T cells [1], but also of natural killer cells [11], of monocytes [12, 13] and of smooth muscle cells (SMC) [1]. Despite these compelling findings described in the literature, in atherosclerosis CXCL1 has been less completely characterized. In human atherosclerosis, but not in healthy arteries, endothelial cells, smooth muscle cells, and

2 2 International Journal of Inflammation macrophages express CXCL1 [15]. Functional observations in CXCL1 gene knockout mice indicated that CXCL1 exerts pro-atherogenic effects, probably related to the specific recruitment and retention of activated Th1 cells and downregulation of a regulatory T-cell response [16]. While these findings indicate that CXCL1 plays a role in atherogenesis, its effect on plaque composition is unclear. Based on our previous observations on gene expression in the mouse model, we hypothesized that CXCL1 mediates development of vulnerable atherosclerotic plaque. Hence, we assessed whether CXCL1 plays a critical role in unstable plaque development by evaluating the effect of antibody-mediated functional inhibition of CXCL1 on lesion phenotype in our vulnerable plaque mouse model. To provide evidence that the findings in the mouse bear relevance for human atherosclerosis, we also investigated the association between CXCL1 concentrations and plaque composition in human carotid endarterectomy lesions. 2. Materials and Methods 2.1. Animals and Surgical Procedures. Apolipoprotein-E deficient mice, 15 2 weeks of age, were obtained from the Jackson Laboratories (Bar Harbor, ME, USA). They were fed a high fat, high cholesterol diet consisting of 15% (wt/wt) cocoa butter and.25% cholesterol (wt/wt) (Diet W, Hope Farms, Woerden, The Netherlands). Two weeks after this diet was initiated, the mice were instrumented with a shear stressaltering device around the right common carotid artery under 2% isoflurane anesthesia, as described previously [2]. Briefly, this device gradually narrows the vessel lumen to 7% of its original diameter. As a consequence, shear stress is lowered upstream of the device, gradually increases inside the device, and is oscillating just downstream of the device [17]. After surgery, the mice were kept on the high fat, high cholesterol diet for the remainder of the experiment. The device remained in situ for 9 weeks. Intervention and control groups consisted of at least 1 animals per group. All animal experiments were approved by the institutional animal ethical committee and performed in compliance with institutional and national guidelines CXCL1 Inhibition. Mice were intravenously injected three times a week in week 2 after surgery with an antimouse bioactivity-neutralizing monoclonal CXCL1 antibody (MAB66, R&D Systems, Abingdon, United Kingdom) in a dose of 25 μg in.2 ml sterile PBS. Untreated castinstrumented mice served as control Isolation and Analysis of Murine Tissues. Following sacrifice, blood was collected and mice were flushed systemically with PBS. Next, the common carotid artery was isolated, embedded in OCT compound (Tissue-Tek, Sakura, Japan), and snap-frozen in liquid nitrogen. For histology, 8μm cryosections were cut and stained for routine histology (Hematoxylin-Eosin, HE), lipids (Oil Red O), collagen (Picrosirius Red), macrophages (anti-cd68, AbD Serotec, Oxford, UK), macrophage activation (anti-major Histocompatibility Complex (MHC) class II, clone M5/11, ATCC TIB-12), and smooth muscle cells (anti-smc α- actin, Sigma-Aldrich, The Netherlands). Subsequently, highresolution images were taken using an Olympus BX- microscope or a Zeiss LSM5 Meta confocal microscope (Zeiss Jena, Germany). Collagen stainings were assessed using crossed circular polarization filters. To calculate lesion size and cellular content relative to the plaque area, all images were analyzed after digitalization by automated image analysis software (Clemex Technologies Inc, Canada) applying thresholding of the stained areas. Necrotic cores (defined as hypocellular plaque cavities devoid of collagen, containing cholesterol clefts) were assessed based on HE and Picrosirius Red stained sections and measured relative to the lesion area. Total cholesterol levels were measured in serum samples by an enzymatic colorimetric method (Cholesterol E, Wako Diagnostics, USA). 2.. Endarterectomy Procedures. All patients in this study were participants of a prospective study aimed at investigating the predictive value of plaque characteristics for longterm outcome (Athero-Express), of which the study methods have been described extensively before [18]. In short, all patients underwent carotid endarterectomy, during which arterial wall tissue was obtained. Next, the culprit lesion, defined as the segment with the largest plaque burden, was processed for histological examination. In addition to HE staining, sections were stained for collagen (Picrosirius Red), SMC (α-actin), and macrophage (CD68) content. Histological assessment was performed in a semiquantitative fashion (none, minor, moderate, heavy), on basis of two independent observers. When results contradicted, a third independent observer was consulted. Subsequently, plaques were designated as fibrous (lipids constitute <1% of lesion surface area), fibroatheromatous (1 % lipids), or atheromatous (>% lipids) based on collagen and HE stainings. Besides, plaques were also designated as SMC dominant or macrophage dominant based on the prevailing cell type obtained from the specific stainings. A 5 mm segment directly adjacent to the culprit lesion was crushed in liquid nitrogen, and subsequently total protein was extracted using 1 ml of TriPure Isolation Reagent (Boehringer Mannheim, Germany), according to the manufacturer s protocol. Then, CXCL1 concentration was measured by Enzyme-Linked Immuno Sorbent Assay (ELISA) according to manufacturers instructions (Quantikine human CXCL1 immunoassay, R&D Systems, Abingdon, United Kingdom) and blinded from histological and clinical data Statistical Analysis. SPSS (version 15., SPSS Inc., USA) was used for all analyses. To test the differences between the treated and the control mice at 9 weeks of cast placement, a student s t-test was used. When these data had a non-gaussian distribution, a nonparametric t-test was used. For the endarterectomy samples, to compare CXCL1 levels between categorical baseline and plaque morphology variables, Kruskal-Wallis and Mann-Whitney tests for non- Gaussian distributed data were used. To investigate the differences between continuous baseline variables and CXCL1,

3 International Journal of Inflammation 3 patients were categorized into quartiles, named 1 through based on increasing CXCL1 concentration. Subsequently, these quartiles were used to sort all other variables. Oneway ANOVA was used to compare continuous variables. The quartiles were also used to visualize the relationship between CXCL1 concentration and plaque morphology. Data are described as mean ± SD or median (Inter Range) as appropriate. A P value of <.5 was considered statistically significant. The authors had full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written. 3. Results 3.1. Suppression of CXCL1 Bioactivity Inhibits Experimental Vulnerable Plaque Formation. As we previously found that CXCL1 is expressed specifically in developing unstable lesions [3], we investigated if CXCL1 is also functionally involved in the development of plaque vulnerability using a mouse model that we developed before [2]. ApoE-deficient mice, in which a shear stress-altering device was applied, were injected with a bioactivity-neutralizing antibody during the onset of plaque formation. As expected, serum total cholesterol levels did not differ between treated and control mice (3.13 ±.6 versus ± 6.7 mmol/l, P =.92). Short-term inhibition of CXCL1 did not influence the extent of plaque development, since we found no difference in lesion size between the treated and the control mice after 9 weeks of shear stress alteration. Because macrophage foam cells are characteristic of atherosclerosis, we measured both plaque lipid (31.3 ± 8.% treated versus 29.5 ± 7.% control) and macrophage content (31.7 ± 7.6% treated versus 27.8 ± 7. control; Figure 1(a)), where both remained unchanged upon CXCL1 inhibition. To assess plaque vulnerability, we determined the amount of collagen in the lesions, which is the main stabilizing component of the plaque. Interestingly, we found a 57% increase in the relative amount of collagen in the plaques following CXCL1 suppression (17.8 ± 6.5% versus 11.3 ± 5.5%, P =.2; Figure 1(b)). The amount of plaque collagen is essentially the result of a balance between collagen deposition and breakdown. Therefore, the increase in collagen may be the result of decreased breakdown predominantly by proteinases secreted by activated macrophages. To determine the extent of immune activation, we measured MHC class II by immunohistochemistry. The cellular morphology, location in the plaque and spatial association of MHC class II staining with macrophage staining by CD68 antibodies in adjacent sections (Figures 1(a) and 1(c)) strongly suggests that MHCII-positive cells are the prime cells expressing this activation marker. We found a 5% reduction in the plaque MHC class II levels following CXCL1 inhibition (6.3 ± 3.3% versus 12.6 ± 7.%, P =.5; Figure 1(c)). In addition, the amount of SMC, which is known to produce collagen, nearly doubled in the CXCL1-suppressed group (13.5 ± 8.% versus 6.3 ± 7.%, P =.3; Figure 1(d)), suggesting that the differences in collagen content may be explained by several factors. The necrotic core is a hallmark component of the vulnerable plaque. To test whether CXCL1 inhibition reduces necrotic core formation, we analyzed both the number of necrotic cores in the lesions as well as their relative size. We found that CXCL1 inhibition resulted in fewer necrotic cores: 38.9 ± 22.1% versus 57.7 ± 2% of the sections covering the entire lesion that contained a necrotic core (P =.2). Moreover, also the relative size of the necrotic cores decreased following antibody treatment from 26. ± 11.% to 15.6 ± 6.1% of the plaque surface area (P =.1; Figure 1(e)) Patient Characteristics. For this study endarterectomy specimens of 16 patients were analyzed. An overview of the patient characteristics is provided in Table 1.Histological examples of the lesions are shown in a previous publication by Verhoeven et al. [18]. The CXCL1 concentration in the specimens ranged from undetectable to 38.8 pg/ml, with a median (interquartile range) of 38.3 pg/ml (1 39 pg/ml). To compare continuous CXCL1 levels to the categorical variables, patients were categorized into quartiles (Figure 2). The variables were then tested for changes across the quartiles. No differences were found comparing risk factors for atherosclerotic disease. The use of medication did not differ significantly between the quartiles High CXCL1 Concentrations Identify Patients with a More Vulnerable Plaque Phenotype. Several significant associations were found between plaque composition and CXCL1 levels. Atheromatous lesions were more prominent at higher CXCL1 concentrations than those classified as fibrous (Rank 53 versus 5, P =.3). Also, higher CXCL1 concentrations were associated with more macrophagedominant plaques (Rank 56 versus 1, P =.9), fewer smooth muscle cells (Rank 39 versus 7, P =.1), and less collagen (Rank 37 versus 63, P =.3). Subsequently, we analyzed the association between plaque composition and CXCL1 levels, by dividing patients into quartiles based upon CXCL1 concentration. More atheromatous lesions were represented in the higher quartiles, while fibrous lesions were more frequently present in the lower quartiles (Figure 3(a)). Macrophage-dominant lesions were mostly found in the top quartiles of CXCL1 content, while SMC dominance associated with lower CXCL1 concentrations (Figure 3(b)). However, no differences were seen in the macrophage staining between the quartiles (Figure (a)). This paradox was the result of an inverse relation between smooth muscle cell staining and increasing CXCL1 concentrations (Figure (b)). Additionally, more collagen staining was seen in the plaques in the lower quartiles (Figure (c)).. Discussion This study may be summarized by its two main findings. First, we showed a functional role of CXCL1 in vulnerable plaque formation in an experimental mouse model for unstable atherosclerotic disease. We observed that inhibiting

4 International Journal of Inflammation α-cxcl1 treatment CD68 area (intima (%)) Control ns Control Anti-CXCL1 Collagen area (intima (%)) (a) Control Anti-CXCL1 MHC-II area (intima (%)) SMC α-actin area (intima (%)) (b) Control Anti-CXCL1 (c) Control Anti-CXCL1 Necrotic core area (intima (%)) (d) Control Anti-CXCL1 (e) Figure 1: Anti-CXCL1 treatment in atherosclerosis susceptible mice results in a change into a more stable lesion phenotype. A flow-altering device around the common carotid artery induced atherosclerosis in ApoE / mice. From week 1 to of lesion development, a bioactivityneutralizing anti-cxcl1 antibody was injected. After 9 weeks, lesions were compared to untreated controls by histology. The pictures show representative histological sections of treated and control mice. All photographs have been made with the same magnification (1x). Scale bars are provided in (e) and represent 1 μm. Data in bar diagrams are the mean values ± SD of at least 8 sections from at least 1 different animals per group. CXCL1 inhibition resulted in a more stable morphology evidenced by unchanged amounts of lesion macrophages (a), increased amounts of collagen (b), decreased macrophage activation (c), increased numbers of SMC (d), and reduced necrotic core size (e). P <.5, P <.1. MHC-II: Major Histocompatibility Complex Class II, SMC: smooth muscle cell.

5 International Journal of Inflammation 5 Clinical characteristics Table 1: Patient baseline measurements for carotid endarterectomy specimens. Total of CXCL1 concentration P value CXCL1 pg/ml 38.3 (1 ) Age (years) 68. ± ± ± ± ± Male (%) Hypertension (%) Smoker (%) Symptomatic disease (%) Statin (%) β blocker (%) ACE inhibitor (%) Serum values Total cholesterol (mmol/l).9 ± ± ± ± ± HDL (mmol/l) 1.2 ±. 1.2 ±. 1. ± ± ±..156 LDL (mmol/l) 3. ± ± ± ± ± Triglycerides (mmol/l) 1.9 ± ± ±.8 2. ± ±.7.3 ApoB (mmol/l).9 ±.2.9 ±.2 1. ±.2 1. ±.3.9 ± Glucose (mmol/l) 6.6 ± ±.7 7. ± ± ± CRP (mg/ml) ( ) (1. 5.8) ( ) ( ) ( ) Baseline data of 16 patients were stratified into four quartiles according to plaque CXCL1 concentration. Values are represented as mean ± SD or median (Inter Range) as appropriate. P values based on ANOVA test between CXCL1 quartiles for continuous variables and nonparametric Mann- Whitney test for categorical variables. CXCL1 with a bioactivity-neutralizing antibody resulted in plaques with less macrophage activation, with more SMC and a subsequent increase in collagen content compared to controls. A decreased number and size of necrotic cores in the lesions of the treated animals further substantiated these findings. This resulted in a lesion phenotype with increased intrinsic stability. Second, we found that in human carotid plaques, CXCL1 levels are associated with lesion morphology. This is evidenced by an increased number of atheromatous plaques with increasing CXCL1 concentrations, and by the association with unstable plaque characteristics, such as macrophage dominance and reduced presence of smooth muscle cells and collagen. Previous studies have shown that inhibition of CXCL1 by antibodies in an intestinal inflammation animal model is an effective way to inhibit its function [19, 2]. The findings in those studies have been ascribed to the reduced amount of Th1 cells that were recruited to the diseased sites and either reduced production of inflammatory cytokines [19] or resulted in inflammatory disease exacerbation [2], depending on the role of T cells in the pathogenesis of the disease. The lesion-stabilizing effect we observed by inhibiting CXCL1 might be explained by a similar mechanism as described above, as it has been shown before that inhibition of the CXCL1-CXCR3 pathway decreases the amount of lesional T cells [16, 21]. Veillard et al. reported that genetic deletion of CXCR3 in ApoE knockout mice reduced early lesion formation, which correlated with an increase in lesional FoxP3-positive regulatory T cells [21]. Recently, it has been reported that genetic deletion of CXCL1 in atherosclerosis susceptible mice also resulted in smaller lesions, which contained fewer CD+ T cells [16]. Surprisingly, despite the decrease in overall T-cell presence in the lesions, a higher expression of FoxP3 as well as of the anti-inflammatory cytokines IL-1 and TGF-β mrna was found. This suggests a more prominent role for regulatory T cells in this model. A similar increase in FoxP3 and TGF-β expression was found in a study using a peptide antagonizing CXCR3 in LDL receptor knockout mice, resulting in smaller lesions in the aortic root and the aorta [22]. A general limitation of our experimental mouse model is that there are few T cells in the induced lesions after 9 weeks of cast placement, approximately 2 per lesion cross section [3].Asaconsequence,wehavebeenunable to find a difference in T-cell numbers following CXCL1 inhibition (data not shown). Nevertheless, the reduction in T cells might have taken place during the early phase of actual inhibition. We performed measurements at only one time point, while the presence of T cells in a lesion changes over time [23]. Also, the effect of small numbers of T cells might provetobeverystrong[2]. The small numbers of T cells do not allow to produce sufficiently reliable data to enable making a comparison with the situation in mice that are deficient in CXCL 1 by gene targeting. This cannot be solved by mrna analyses, as there was no RNA available from the endarterectomy samples that have been used. Besides facilitating recruitment of leukocytes, CXCL1 also induces and sustains a dominant

6 6 International Journal of Inflammation P =.1 Plaque concentration of CXCL1 (pg/ml) Percent Figure 2: Distribution of CXCL1 measurements from each patient across quartiles. Atherosclerotic plaques were obtained from 16 patients during carotid endarterectomy. In the plaque segment directly adjacent to the culprit lesion, the content of the chemokine CXCL1 was measured by ELISA. Based on these measurements, patients were divided into one of four quartiles according to the CXCL1 concentration. 1 represents the lesions with the lowest CXCL1 values, whereas quartile contains the highest. This figure shows the individual measurements for all patients and the distribution across each quartile of plaque CXCL1. Horizontal bars represent the median concentration of each quartile Atheromatous Fibroatheromatous Fibrous (a) P =.9 Th1-type response by increasing the production of IFN-γ, providing a positive feedback loop [25]. Neutralization of CXCL1 might therefore result in a decreased concentration of IFN-γ in the lesions, which is considered a key cytokine in disease progression [26] and macrophage activation [27]. This notion is in agreement with our findings regarding macrophage activation status as measured by MHC class II expression [27], SMC presence [27], and collagen accumulation [28]. An alternative or additional mechanism of plaque modification by CXCL1 inhibition might be operative as well. Not only Th1 and NK cells, as major IFN-γ producers, are responsive to CXCL1, but also plasmacytoid dendritic cells (PDC) express CXCR3 and are recruited via this receptor [29]. The numbers of circulating PDC are reduced in human atherosclerosis, while immunohistochemical analysis has indicated that they are recruited to advanced plaques [3, 31]. PDC are known to produce large quantities of IFN-α upon activation, and this cytokine functions as an inflammatory amplifier in the atherosclerotic plaque [32]. Additionally, a minor subset of monocytes has been shown to be recruited via CXCR3 to inflammatory environments [7] and might thus contribute to plaque instability as well. Further research is needed to understand the mechanisms underlying the effects of CXCL1 inhibition. For example, apoptosis could have an important role in the observed differences in plaque composition and size. It could be of interest to study the role of SMC in more detail by staining serial sections for proliferation and apoptosis on the one hand and alpha actin on the other hand. In addition, in situ zymography studies could be performed to get a clearer idea of the possible role of collagenolytic Percent 5 25 SMC dominant Mo dominant (b) Figure 3: Plaque concentrations of CXCL1 are associated with lesion morphology. Individual values were divided over quartiles (see Figure 1). By using ANOVA, we tested to see if plaque CXCL1 concentrations associated with lesion morphology. Increasing plaque concentrations of CXCL1 were indeed associated with an increase in atheromatous-type lesions (a) and a decrease of SMC dominant lesions and an increase of macrophage dominant lesions (b). activity in the lesions. It would be very interesting to know more about the role of the proinflammatory cytokine microenvironment in the plaque in the experiments with CXCL 1 inhibition. From the present data, we cannot exclude that by inhibiting CXCL1 we just postponed the process of the lesion formation and maturation rather than permanently altering it. For this argument and because of the ongoing discussion on the quality of animal models representing human disease states, we also performed a parallel study in humans.

7 International Journal of Inflammation 7 Macrophages (weighed staining) SMC α-actin (weighed staining) P = ns (a) P =.1 (b) Mach et al. [15] were the first to associate CXCL1 with human atherosclerosis, reporting that CXCL1 was expressed in several stages of the disease. CXCL1 was shown to be present on the surface of endothelial cells, suggesting a role in the recruitment and adhesion of CXCR3 positive cells from the circulation. They also showed broad expression of CXCL1 by SMC and macrophages throughout the lesion. Little is known about the effects of CXCL1 on human atherosclerosis development and clinical outcome. Previous studies sought to correlate plasma CXCL1 levels with the occurrence of coronary heart disease (CHD). In a casecontrol study it was found that CHD risk was associated with an increase in serum CXCL1 [33]. A later prospective study showed that indeed increased levels of CXCL1 exist prior to CHD, but was not considered an independent risk factor [3]. Our study is the first one to correlate lesional CXCL1 protein content with plaque morphology in humans and to be a possible predictor of plaque vulnerability. Obviously, this cannot be measured directly in living patients, so it would be of interest to know the correlation between lesional and systemic levels of CXCL1. There were no plasma samples available to measure the systemic CXCL 1 levels and correlate them to the lesional levels, so this has to await future work. In conclusion, we showed that the chemokine CXCL1 plays a functional role in the destabilization of atherosclerotic plaques in mice and is specifically upregulated in vulnerable plaques in humans. Since the experimental data were paralleled by similar findings in human unstable lesions, CXCL1 might be regarded as a new lead into understanding the process of plaque destabilization. Collagen (weighed staining) 3 2 P =.3 Acknowledgments The authors would like to thank Rien van Haperen (Erasmus University Medical Center Rotterdam, The Netherlands) and Ben van Middelaar (University Medical Center Utrecht, The Netherlands) for their help with the animal studies. This research was supported by the Netherlands Heart Foundation Grant 22T5. 1 (c) Figure : Increasing carotid plaque CXCL1 concentrations are associated with a less fibrous phenotype. Plaque concentrations of CXCL1 were associated with histological qualities of the lesion by ANOVA. Staining for macrophages, SMC, or collagen was quantified as none, minor, moderate, or high. Based on these qualifications a weighed value for average staining intensity was calculated (1, none; 2, minor; 3, moderate;, heavy staining). Macrophage staining was not associated with plaque CXCL1 (a), whereas high plaque concentrations of CXCL1 were associated with decreasing amounts of smooth muscle cells (SMC) (b), and decreasing collagen (c). References [1] R. Virmani, F. D. Kolodgie, A. P. Burke, A. Farb, and S. M. Schwartz, Lessons from sudden coronary death: a comprehensive morphological classification scheme for atherosclerotic lesions, Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 2, no. 5, pp , 2. [2] C. Cheng, D. Tempel, R. van Haperen et al., Atherosclerotic lesion size and vulnerability are determined by patterns of fluid shear stress, Circulation, vol. 113, no. 23, pp , 26. [3] C. Cheng, D. Tempel, R. van Haperen et al., Shear stressinduced changes in atherosclerotic plaque composition are modulated by chemokines, Journal of Clinical Investigation, vol. 117, no. 3, pp , 27. [] A. D. Luster and J. V. Ravetch, Biochemical characterization of a γ interferon-inducible cytokine (IP-1), Journal of Experimental Medicine, vol. 166, no., pp , 1987.

8 8 International Journal of Inflammation [5] M. Inngjerdingen, B. Damaj, and A. A. Maghazachi, Expression and regulation of chemokine receptors in human natural killer cells, Blood, vol. 97, no. 2, pp , 21. [6] M.Loetscher,P.Loetscher,N.Brass,E.Meese,andB.Moser, Lymphocyte-specific chemokine receptor CXCR3: regulation, chemokine binding and gene localization, European Journal of Immunology, vol. 28, no. 11, pp , [7] M. J. Janatpour, S. Hudak, M. Sathe, J. D. Sedgwick, and L. M. McEvoy, Tumor necrosis factor-dependent segmental control of MIG expression by high endothelial venules in inflamed lymph nodes regulates monocyte recruitment, Journal of Experimental Medicine, vol. 19, no. 9, pp , 21. [8] J. G. Hyun, G. Lee, J. B. Brown et al., Anti-interferoninducible chemokine, CXCL1, reduces colitis by impairing T helper-1 induction and recruitment in mice, Inflammatory Bowel Diseases, vol. 11, no. 9, pp , 25. [9] B. T. Fife, K. J. Kennedy, M. C. Paniagua et al., CXL1 (IFN-γ-inducible protein-1) control of encephalitogenic CD+ T cell accumulation in the central nervous system during experimental autoimmune encephalomyelitis, Journal of Immunology, vol. 166, no. 12, pp , 21. [1] M. Loetscher, B. Gerber, P. Loetscher et al., Chemokine receptor specific for IP1 and Mig: structure, function, and expression in activate T-Lymphocytes, Journal of Experimental Medicine, vol. 18, no. 3, pp , [11] A. D. Luster, Mechanisms of disease: chemokines chemotactic cytokines that mediate inflammation, New England Journal of Medicine, vol. 338, no. 7, pp. 36 5, [12] D. D. Taub, D. L. Longo, and W. J. Murphy, Human interferon-inducible protein-1 induces mononuclear cell infiltration in mice and promotes the migration of human T lymphocytes into the peripheral tissues of human peripheral blood lymphocytes-scid mice, Blood, vol. 87, no., pp , [13] D. D. Taub, A. R. Lloyd, K. Conlon et al., Recombinant human interferon-inducible protein 1 is a chemoattractant for human monocytes and T lymphocytes and promotes T cell adhesion to endothelial cells, Journal of Experimental Medicine, vol. 177, no. 6, pp , [1] X. Wang, T. L. Yue, E. H. Ohlstein, C. P. Sung, and G. Z. Feuerstein, Interferon-inducible protein-1 involves vascular smooth muscle cell migration, proliferation, and inflammatory response, Journal of Biological Chemistry, vol. 271, no. 39, pp , [15] F. Mach, A. Sauty, A. S. Iarossi et al., Differential expression of three t lymphocyte-activating CXC chemokines by human atheroma-associated cells, Journal of Clinical Investigation, vol. 1, no. 8, pp , [16] E.A.Heller,E.Liu,A.M.Tageretal., ChemokineCXCL1 promotes atherogenesis by modulating the local balance of effector and regulatory T cells, Circulation, vol. 113, no. 19, pp , 26. [17] C. Cheng, R. van Haperen, M. de Waard et al., Shear stress affects the intracellular distribution of enos: direct demonstration by a novel in vivo technique, Blood, vol. 16, no. 12, pp , 25. [18] B. A. N. Verhoeven, E. Velema, A. H. Schoneveld et al., Athero-express: differential atherosclerotic plaque expression of mrna and protein in relation to cardiovascular events and patient characteristics. Rationale and design, European Journal of Epidemiology, vol. 19, no. 12, pp , 2. [19] U. P. Singh, S. Singh, D. D. Taub, and J. W. Lillard, Inhibition of IFN-γ-inducible protein-1 abrogates colitis in IL-1-/- mice, Journal of Immunology, vol. 171, no. 3, pp , 23. [2] C. Nakajima, T. Mukai, N. Yamaguchi et al., Neutralization of IFN-inducible protein 1/CXCL1 exacerbates experimental autoimmune encephalomyelitis, European Journal of Immunology, vol. 32, no. 6, pp , 22. [21] N. R. Veillard, S. Steffens, G. Pelli et al., Differential influence of chemokine receptors CCR2 and CXCR3 in development of atherosclerosis in vivo, Circulation, vol. 112, no. 6, pp , 25. [22] E. J. A. van Wanrooij, S. C. A. de Jager, T. van Es et al., CXCR3 antagonist NBI-733 attenuates atherosclerotic plaque formation in LDL receptor-deficient mice, Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 28, no. 2, pp , 28. [23] J. Khallou-Laschet, G. Caligiuri, E. Groyer et al., The proatherogenic role of T cells requires cell division and is dependent on the stage of the disease, Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 26, no. 2, pp , 26. [2] U. Schleicher, A. Hesse, and C. Bogdan, Minute numbers of contaminant CD8+ T cells or CD11b +CD11c+ NK cells are the source of IFN-γ in IL-12/IL-18-stimulated mouse macrophage populations, Blood, vol. 15, no. 3, pp , 25. [25] V. Gangur, F. E.R. Simons, and K. T. Hayglass, Human IP- 1 selectively promotes dominance of polyclonally activated and environmental antigendriven IFN-γ over IL- responses, FASEB Journal, vol. 12, no. 9, pp , [26] E.J.Harvey andd.p.ramji, Interferon-γ and atherosclerosis: pro- or anti-atherogenic? Cardiovascular Research, vol. 67, no. 1, pp. 11 2, 25. [27] K. Schroder, P. J. Hertzog, T. Ravasi, and D. A. Hume, Interferon-γ: an overview of signals, mechanisms and functions, Journal of Leukocyte Biology, vol. 75, no. 2, pp , 2. [28] E. P. Amento, N. Ehsani, H. Palmer, and P. Libby, Cytokines and growth factors positively and negatively regulate interstitial collagen gene expression in human vascular smooth muscle cells, Arteriosclerosis and Thrombosis, vol. 11, no. 5, pp , [29] M. Cella, D. Jarrossay, F. Faccheth et al., Plasmacytoid monocytes migrate to inflamed lymph nodes and produce large amounts of type I interferon, Nature Medicine, vol. 5, no. 8, pp , [3] E. A. Van Vré, J. M. Bosmans, I. Van Brussel et al., Immunohistochemical characterisation of dendritic cells in human atherosclerotic lesions: possible pitfalls, Pathology, vol. 3, no. 3, pp , 211. [31] E. A. Van Vré, V. Y. Hoymans, H. Bult et al., Decreased number of circulating plasmacytoid dendritic cells in patients with atherosclerotic coronary artery disease, Coronary Artery Disease, vol. 17, no. 3, pp , 26. [32] A. Niessner, M. S. Shin, O. Pryshchep, J. J. Goronzy, E. L. Chaikof, and C. M. Weyand, Synergistic proinflammatory effects of the antiviral cytokine interferon-α and toll-like receptor ligands in the atherosclerotic plaque, Circulation, vol. 116, no. 18, pp , 27. [33] D. Rothenbacher, S. Müller-Scholze, C. Herder, W. Koenig, and H. Kolb, Differential expression of chemokines, risk of

9 International Journal of Inflammation 9 stable coronary heart disease, and correlation with established cardiovascular risk markers, Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 26, no. 1, pp , 26. [3] C. Herder, J. Baumert, B. Thorand et al., Chemokines and incident coronary heart disease: results from the MON- ICA/KORA Augsburg case-cohort study, , Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 26, no. 9, pp , 26.